NFV 1112.08. SPX 1112.97. Convergence is happening fast again. Note that if we sold the open yesterday at ~1120 SPX, adding to a losing position, we would have had a decent gain on two units on convergence today. There may be an optimal algorithm on dynamically contracting and expanding 7->14->24[->35->55->70] add points (empirically derived). I would not go any tighter, but knowing when to expand it and then contracting it again from an expansion (less aggressive, more aggressive, respectively) could raise Sortino and Sharp ratios considerably.
Where: This DR is what we are trying to dynamically manage by expanding add points when momentum is high (we are trying to guess what the real f(x) is in DR). http://en.wikipedia.org/wiki/Sortino_ratio
Can you tell what the risk adjusted return is in advance, in other words strip away the formulas, no one know the return or variance in advance, why not take the trade at es 1050 with a 100 point stop.
I am not sure I follow you. RAR is based on your "drawdown", and you won't know what it is until you enter a trade and decide to add or not, and then where the market goes and when you decide to exit. That is why f(x) above in the DR formula, or the pdf of the returns, is so important. If the pdf is telling you that the odds of a +/-70 divergence from NFV has quintupled (a very high momentum environment), maybe we change our add points to be more conservative. See? Huh?
I guess what I am asking without going into the formulas is why take the trade when there is little divergence, when a few weeks ago there was massive divergence. I am asking rhetorically how much confidence you have in your system, my point about the risk adjusted return is that you won't really know your drawdown or return until you enter/exit the trade. Hope that makes sense.
Well, that is the point of Taleb's "Black Swan" - most traders calibrate their models to what they have seen, and probably incorrectly assume that they have a good idea of what the true pdf is from this data. I assumed that when I saw the flash crash, that 40 was a colossal number of divergence (although that was +40 and there may be an asymmetry between relative cheapness and relative expensive), so I set out to structure my money management add points around this boundary condition. Once I saw -70, obviously this theory is wrong. But here is the rub, if you wait for +/-70 again, you may be waiting a very long time to put on a trade. In this case, I hope you have another means of income. Trading is like trying to cross a busy multi-lane highway with cars coming from both sides, trying to pick up money that fell scattered from somebodies briefcase. You sort of time how fast you have to go not to get run over by the next car (the pdf) as you make progress crossing the highway picking up $20s, $100s, etc, and you hope that cars stay in their lanes more or less. Every once in a while cars do very unexpected things and you get run over. Why cross the busy highway at all, or why not wait until the highway settles down? Why did the chicken cross the road? To get to the other side. I have very high confidence in NFV at this point. I have less confidence on my ability to guess the (true) pdf better than anybody else. Hence my add points (7->14->24->etc) are where if you wanted to improve on this system, is where you can get the biggest benefits for yourself. Shortie claims he does better than me with his add points, but he has never stated what they are or why he used them.